Exploring and Analyizing Wine Enthusiast Reviews

Prerequisites

Load required packages

library(tidyverse)
library(dplyr)
library(ggplot2)
library(rtweet)
library(readr)
library(DataExplorer)

Dataset

Import processed data, which can be found here.

#read preprocessed data
wines <- read.csv(file = '../data/processed_data/wines.csv')

Get sample of dataset

#set seed value to birthday of Ricardo Rodriguez, American wrestler and ring announcer and Dr. Reinaldo (Rei) Sanchez-Arias
set.seed(19630217)

#set percentage to test with for simplicity, if needed
percentage <- 5
wine_sample<- sample_n(wines, percentage/100*nrow(wines))

Split Taster data into different Data Frame

tasters <- wines %>%
  select(taster_name, taster_twitter_handle) %>% unique()
tasters

Drop taster_twitter_handle in wines dataframe

wines <- wines %>%
  select(-taster_twitter_handle)
head(wines)

Add Reviewer profile info

Each reviewer has there own bias. To offset that we made a “profile” for each reviewer which includes characteristics like: avg_points, sd_points, and var_points

taster_rating_profile <- wines %>%
  group_by(taster_name) %>%
  summarize(
    avg_points = mean(points),
    sd_points = sd(points),
    var_points = var(points),
    reviews = n()
  )

tasters <- inner_join(tasters, taster_rating_profile, by = "taster_name")
head(tasters)

Add Rating Classification

Add following classification to wine dataset as found on the website:

Category	Rating	Description
Classic	98-100	The pinnacle of quality.
Superb	94-97	A great achievement.
Excellent	90-93	Highly recommended.
Very Good	87-89	Often good value; well recommended.
Good	83-86	Suitable for everyday consumption; often good value.
Acceptable	80-82	Can be employed in casual, less-critical circumstances

# function to add rating
rating_category <- function(points){
  if(points>=98){
    return("Classic")
  }
  else if (points>=94){
    return("Superb")
  }
  else if(points>=90){
    return("Excellent")
  }
  else if(points>=87){
    return("Very Good")
  }
  else if(points>=83){
    return("Good")
  }
  else{
    return("Acceptable")
  }
}

wines<- wines %>%
  rowwise() %>%
  mutate(rating_category = rating_category(points))
head(wines)

Add Adjusted Points

Since, each reviewer has a different bias we created a normalized metric, norm_points, by looking at the number of standard deviatioins a wine is from the reviewer’s avg_points. This gives use a more accurate representation of which which wines are better than the rest.

normalize_points <- function(data){
  left_join(data, tasters, by = "taster_name")%>%
    rowwise() %>%
    mutate(norm_points = (points-avg_points)/sd_points) %>%
    select(-avg_points, -sd_points, -var_points, -taster_twitter_handle, -reviews)
}

wines <- normalize_points(wines)
head(wines) 

Data Sanitation

Vintage seems to have year 7200

wines <- wines %>%
  filter(vintage<2019)

Data Exploration

Univariate Exploration

Correlation price by points, using DataExplorer library which can be found here

# TODO: IZZY

Alcohol Amount

# TODO: IZZY

Category

# TODO: IZZY

Vintage

Count wines per year (Note: Data has been sanitized)

wines %>%
  group_by(vintage) %>%
  summarize(count = n())

Grouping rowwise data frame strips rowwise nature

wines %>%
  ggplot() +
  geom_bar(mapping = aes(x=vintage))

Winery

Count the number of wines per winery (in a column graph)

wines %>%
  group_by(winery) %>%
  summarize(count = n()) %>%
  arrange(desc(count)) %>%
  slice(1:15) %>%
  ggplot() +
  geom_col(mapping = aes(x=count, y =winery)) 

Province

Count the number of wines per province (in a column graph)

wines %>% 
  group_by(province) %>% 

Price

(graphs to understand where the majority of the data is. (in a column graph))

mean_price <- (mean(wines$price, na.rm = TRUE))
sd_price <- (sd(wines$price, na.rm = TRUE))
min_price <- (min(wines$price, na.rm = TRUE))
max_price <- print(max(wines$price, na.rm = TRUE))

[1] 3300

ggplot(mapping = aes(mean_price, sd_price, min_price, max_price))+
  geom_boxplot()

Duplicated aesthetics after name standardisation: 

Points

(graphs to understand where the majority of the data is. (in a column graph))

print(mean(wines$points))
print(sd(wines$points))
print(min(wines$points))
print(max(wines$points))

Points distribution by Reviewer

wines %>%
  ggplot() +
  geom_boxplot(aes(y=taster_name, x=points)) +
  geom_vline(xintercept = mean(wines$points))

Multivariate Exploration

Price by Points

Notice the data is “stacked” and the socres range from 80-100

wines %>% 
  ggplot() +
  geom_point(mapping = (aes(x = points, y = price)), na.rm = T, alpha = 0.15) +
  labs(title = "Price by Points", x = "Points", y = "Price")

TODO: IZZY (Why did we log this?)

wines %>% 
  ggplot() +
  geom_point(mapping = (aes(x = points, y = log(price))), na.rm = T, alpha = 0.15) +
  labs(title = "log(Price) by Points", x = "Points", y = "log(Price)")

Data Analysis

#Find the best province for wine using the average points across the 1,000 samples #drop the descriptions or just select price? set points to max(points)

mean_points <- mean(wine_sample$points)
mean_points

best_province <- wine_sample %>% 
  group_by(points) %>% 
  filter(points > mean_points) %>% 
  arrange(desc(points))
best_province

Best wine, by variety

#wine_best_variety <- 
wines %>% 
  group_by(variety) %>% 
  summarise(mean_points = mean(points)) %>% 
  arrange(desc(mean_points)) 
  

user_price <- readline(prompt = "How much are you willing to spend on a bottle?")
user_price <- as.integer(user_price)

wines %>% 
  filter(price <= user_price) %>% 
  arrange(desc(points)) %>% 
  select(title, price, points)

Conclusion